from sklearn_benchmarks.reporting.hp_match import HPMatchReporting

scikit-learn vs. ONNX Runtime

reporting = HPMatchReporting(against_lib="onnx", config="config.yml")
reporting.make_report()

We assume here there is a perfect match between the hyperparameters of both librairies. For a given set of parameters and a given dataset, we compute the speedup time scikit-learn / time onnx. For instance, a speedup of 2 means that onnx is twice as fast as scikit-learn for a given set of parameters and a given dataset.

KNeighborsClassifier_brute_force

onnx (1.10.1) vs. scikit-learn (1.0.dev0)

Speedup barplots

All estimators share the following parameters: algorithm=brute.

WARNING! Mismatch between validation scores for 12 predictions (75.0%). See below.

	estimator	function	n_samples_train	n_samples	n_features	accuracy_score_sklearn	accuracy_score_onnx	diff_accuracy_scores
1	KNeighborsClassifier_brute_force	predict	100000	1000	100	0.663	1.000	0.337
2	KNeighborsClassifier_brute_force	predict	100000	1	100	1.000	0.757	0.243
4	KNeighborsClassifier_brute_force	predict	100000	1000	100	0.757	0.882	0.125
7	KNeighborsClassifier_brute_force	predict	100000	1000	100	0.882	1.000	0.118
8	KNeighborsClassifier_brute_force	predict	100000	1	100	1.000	0.757	0.243
10	KNeighborsClassifier_brute_force	predict	100000	1000	100	0.882	0.663	0.219
13	KNeighborsClassifier_brute_force	predict	100000	1000	100	0.757	1.000	0.243
14	KNeighborsClassifier_brute_force	predict	100000	1	100	1.000	0.922	0.078
16	KNeighborsClassifier_brute_force	predict	100000	1000	100	0.663	0.929	0.266
19	KNeighborsClassifier_brute_force	predict	100000	1000	2	0.896	1.000	0.104
20	KNeighborsClassifier_brute_force	predict	100000	1	2	1.000	0.922	0.078
22	KNeighborsClassifier_brute_force	predict	100000	1000	2	0.922	0.896	0.026

Raw results

	estimator	function	n_samples_train	n_samples	n_features	mean_duration_sklearn	std_duration_sklearn	iteration_throughput	latency	n_jobs	n_neighbors	mean_duration_onnx	std_duration_onnx	accuracy_score_onnx	speedup	std_speedup	sklearn_profiling	onnx_profiling
0	KNeighborsClassifier_brute_force	fit	100000	100000	100	0.011	0.0	7.538	0.0	-1	1	15.479	0.032	0.663	0.001	0.001	See	See
3	KNeighborsClassifier_brute_force	fit	100000	100000	100	0.011	0.0	7.557	0.0	-1	5	0.287	0.005	1.000	0.037	0.037	See	See
6	KNeighborsClassifier_brute_force	fit	100000	100000	100	0.011	0.0	7.571	0.0	1	100	15.460	0.038	0.882	0.001	0.001	See	See
9	KNeighborsClassifier_brute_force	fit	100000	100000	100	0.010	0.0	7.644	0.0	-1	100	0.286	0.005	1.000	0.037	0.037	See	See
12	KNeighborsClassifier_brute_force	fit	100000	100000	100	0.011	0.0	7.590	0.0	1	5	3.381	0.011	0.896	0.003	0.003	See	See
15	KNeighborsClassifier_brute_force	fit	100000	100000	100	0.010	0.0	7.643	0.0	1	1	0.213	0.004	1.000	0.049	0.049	See	See
18	KNeighborsClassifier_brute_force	fit	100000	100000	2	0.004	0.0	0.420	0.0	-1	1	3.762	0.013	0.929	0.001	0.001	See	See
21	KNeighborsClassifier_brute_force	fit	100000	100000	2	0.004	0.0	0.376	0.0	-1	5	0.242	0.004	1.000	0.018	0.018	See	See

	estimator	function	n_samples_train	n_samples	n_features	mean_duration_sklearn	std_duration_sklearn	iteration_throughput	latency	n_jobs	n_neighbors	mean_duration_onnx	std_duration_onnx	accuracy_score_onnx	speedup	std_speedup	sklearn_profiling	onnx_profiling
1	KNeighborsClassifier_brute_force	predict	100000	1000	100	2.070	0.234	0.000	0.002	-1	1	0.283	0.007	1.000	7.314	7.316	See	See
2	KNeighborsClassifier_brute_force	predict	100000	1	100	0.021	0.002	0.000	0.021	-1	1	15.290	0.136	0.757	0.001	0.001	See	See
4	KNeighborsClassifier_brute_force	predict	100000	1000	100	2.721	0.078	0.000	0.003	-1	5	15.330	0.082	0.882	0.178	0.178	See	See
5	KNeighborsClassifier_brute_force	predict	100000	1	100	0.022	0.003	0.000	0.022	-1	5	0.286	0.006	1.000	0.076	0.076	See	See
7	KNeighborsClassifier_brute_force	predict	100000	1000	100	1.906	0.009	0.000	0.002	1	100	0.285	0.005	1.000	6.693	6.694	See	See
8	KNeighborsClassifier_brute_force	predict	100000	1	100	0.019	0.000	0.000	0.019	1	100	15.169	0.053	0.757	0.001	0.001	See	See
10	KNeighborsClassifier_brute_force	predict	100000	1000	100	2.701	0.042	0.000	0.003	-1	100	15.353	0.008	0.663	0.176	0.176	See	See
11	KNeighborsClassifier_brute_force	predict	100000	1	100	0.021	0.002	0.000	0.021	-1	100	0.286	0.005	1.000	0.075	0.075	See	See
13	KNeighborsClassifier_brute_force	predict	100000	1000	100	1.882	0.004	0.000	0.002	1	5	0.241	0.006	1.000	7.795	7.797	See	See
14	KNeighborsClassifier_brute_force	predict	100000	1	100	0.019	0.000	0.000	0.019	1	5	3.724	0.014	0.922	0.005	0.005	See	See
16	KNeighborsClassifier_brute_force	predict	100000	1000	100	1.150	0.004	0.001	0.001	1	1	3.768	0.013	0.929	0.305	0.305	See	See
17	KNeighborsClassifier_brute_force	predict	100000	1	100	0.018	0.000	0.000	0.018	1	1	0.241	0.005	1.000	0.075	0.075	See	See
19	KNeighborsClassifier_brute_force	predict	100000	1000	2	1.783	0.036	0.000	0.002	-1	1	0.240	0.005	1.000	7.420	7.421	See	See
20	KNeighborsClassifier_brute_force	predict	100000	1	2	0.005	0.004	0.000	0.005	-1	1	3.727	0.011	0.922	0.001	0.001	See	See
22	KNeighborsClassifier_brute_force	predict	100000	1000	2	2.628	0.077	0.000	0.003	-1	5	3.775	0.020	0.896	0.696	0.696	See	See
23	KNeighborsClassifier_brute_force	predict	100000	1	2	0.009	0.005	0.000	0.009	-1	5	0.240	0.005	1.000	0.038	0.038	See	See

KNeighborsClassifier_kd_tree

onnx (1.10.1) vs. scikit-learn (1.0.dev0)

Speedup barplots

All estimators share the following parameters: algorithm=kd_tree.

WARNING! Mismatch between validation scores for 12 predictions (75.0%). See below.

	estimator	function	n_samples_train	n_samples	n_features	accuracy_score_sklearn	accuracy_score_onnx	diff_accuracy_scores
1	KNeighborsClassifier_kd_tree	predict	1000000	1000	10	0.929	1.000	0.071
2	KNeighborsClassifier_kd_tree	predict	1000000	1	10	1.000	0.946	0.054
4	KNeighborsClassifier_kd_tree	predict	1000000	1000	10	0.946	0.951	0.005
7	KNeighborsClassifier_kd_tree	predict	1000000	1000	10	0.951	1.000	0.049
8	KNeighborsClassifier_kd_tree	predict	1000000	1	10	1.000	0.946	0.054
10	KNeighborsClassifier_kd_tree	predict	1000000	1000	10	0.951	0.929	0.022
13	KNeighborsClassifier_kd_tree	predict	1000000	1000	10	0.946	1.000	0.054
14	KNeighborsClassifier_kd_tree	predict	1000000	1	10	1.000	0.911	0.089
16	KNeighborsClassifier_kd_tree	predict	1000000	1000	10	0.929	0.894	0.035
19	KNeighborsClassifier_kd_tree	predict	1000	1000	2	0.891	1.000	0.109
20	KNeighborsClassifier_kd_tree	predict	1000	1	2	1.000	0.911	0.089
22	KNeighborsClassifier_kd_tree	predict	1000	1000	2	0.911	0.891	0.020

Raw results

	estimator	function	n_samples_train	n_samples	n_features	mean_duration_sklearn	std_duration_sklearn	iteration_throughput	latency	n_jobs	n_neighbors	mean_duration_onnx	std_duration_onnx	accuracy_score_onnx	speedup	std_speedup	sklearn_profiling	onnx_profiling
0	KNeighborsClassifier_kd_tree	fit	1000000	1000000	10	2.745	0.028	0.029	0.0	-1	1	106.853	0.000	0.929	0.026	0.026	See	See
3	KNeighborsClassifier_kd_tree	fit	1000000	1000000	10	3.295	0.030	0.024	0.0	-1	5	2.311	0.207	1.000	1.426	1.432	See	See
6	KNeighborsClassifier_kd_tree	fit	1000000	1000000	10	3.315	0.052	0.024	0.0	1	100	107.347	0.000	0.951	0.031	0.031	See	See
9	KNeighborsClassifier_kd_tree	fit	1000000	1000000	10	3.316	0.066	0.024	0.0	-1	100	2.618	0.164	1.000	1.267	1.269	See	See
12	KNeighborsClassifier_kd_tree	fit	1000000	1000000	10	3.635	0.073	0.022	0.0	1	5	0.039	0.013	0.891	93.662	99.077	See	See
15	KNeighborsClassifier_kd_tree	fit	1000000	1000000	10	3.705	0.114	0.022	0.0	1	1	0.004	0.000	1.000	824.261	824.541	See	See
18	KNeighborsClassifier_kd_tree	fit	1000	1000	2	0.001	0.001	0.018	0.0	-1	1	0.053	0.001	0.894	0.017	0.017	See	See
21	KNeighborsClassifier_kd_tree	fit	1000	1000	2	0.001	0.000	0.030	0.0	-1	5	0.004	0.000	1.000	0.120	0.120	See	See

	estimator	function	n_samples_train	n_samples	n_features	mean_duration_sklearn	std_duration_sklearn	iteration_throughput	latency	n_jobs	n_neighbors	mean_duration_onnx	std_duration_onnx	accuracy_score_onnx	speedup	std_speedup	sklearn_profiling	onnx_profiling
1	KNeighborsClassifier_kd_tree	predict	1000000	1000	10	0.801	0.921	0.000	0.001	-1	1	2.351	0.286	1.000	0.341	0.343	See	See
2	KNeighborsClassifier_kd_tree	predict	1000000	1	10	0.002	0.001	0.000	0.002	-1	1	105.586	0.000	0.946	0.000	0.000	See	See
4	KNeighborsClassifier_kd_tree	predict	1000000	1000	10	1.043	0.468	0.000	0.001	-1	5	107.361	0.000	0.951	0.010	0.010	See	See
5	KNeighborsClassifier_kd_tree	predict	1000000	1	10	0.003	0.001	0.000	0.003	-1	5	2.309	0.179	1.000	0.001	0.001	See	See
7	KNeighborsClassifier_kd_tree	predict	1000000	1000	10	5.536	0.393	0.000	0.006	1	100	2.577	0.061	1.000	2.148	2.149	See	See
8	KNeighborsClassifier_kd_tree	predict	1000000	1	10	0.003	0.001	0.000	0.003	1	100	107.975	0.000	0.946	0.000	0.000	See	See
10	KNeighborsClassifier_kd_tree	predict	1000000	1000	10	3.100	0.220	0.000	0.003	-1	100	107.164	0.000	0.929	0.029	0.029	See	See
11	KNeighborsClassifier_kd_tree	predict	1000000	1	10	0.005	0.001	0.000	0.005	-1	100	2.312	0.169	1.000	0.002	0.002	See	See
13	KNeighborsClassifier_kd_tree	predict	1000000	1000	10	1.634	0.260	0.000	0.002	1	5	0.005	0.000	1.000	345.865	346.172	See	See
14	KNeighborsClassifier_kd_tree	predict	1000000	1	10	0.002	0.000	0.000	0.002	1	5	0.034	0.001	0.911	0.045	0.045	See	See
16	KNeighborsClassifier_kd_tree	predict	1000000	1000	10	0.881	0.251	0.000	0.001	1	1	0.053	0.001	0.894	16.743	16.745	See	See
17	KNeighborsClassifier_kd_tree	predict	1000000	1	10	0.001	0.000	0.000	0.001	1	1	0.005	0.000	1.000	0.200	0.200	See	See
19	KNeighborsClassifier_kd_tree	predict	1000	1000	2	0.022	0.011	0.001	0.000	-1	1	0.005	0.000	1.000	4.948	4.950	See	See
20	KNeighborsClassifier_kd_tree	predict	1000	1	2	0.002	0.001	0.000	0.002	-1	1	0.034	0.001	0.911	0.070	0.070	See	See
22	KNeighborsClassifier_kd_tree	predict	1000	1000	2	0.021	0.001	0.001	0.000	-1	5	0.034	0.001	0.891	0.616	0.616	See	See
23	KNeighborsClassifier_kd_tree	predict	1000	1	2	0.002	0.000	0.000	0.002	-1	5	0.004	0.000	1.000	0.393	0.393	See	See

HistGradientBoostingClassifier_best

onnx (1.10.1) vs. scikit-learn (1.0.dev0)

Speedup barplots

All estimators share the following parameters: learning_rate=0.01, n_iter_no_change=10.0, max_leaf_nodes=100.0, max_bins=255.0, min_samples_leaf=100.0, max_iter=300.0.

WARNING! Mismatch between validation scores for 1 prediction (100.0%). See below.

	estimator	function	n_samples_train	n_samples	n_features	accuracy_score_sklearn	accuracy_score_onnx	diff_accuracy_scores
1	HistGradientBoostingClassifier_best	predict	100000	1000	100	0.824	1.0	0.176

Raw results

	estimator	function	n_samples_train	n_samples	n_features	mean_duration_sklearn	std_duration_sklearn	n_iter	iteration_throughput	latency	mean_duration_onnx	std_duration_onnx	accuracy_score_onnx	speedup	std_speedup	sklearn_profiling	onnx_profiling
0	HistGradientBoostingClassifier_best	fit	100000	100000	100	100.054	0.0	300	0.001	0.001	0.452	0.02	0.824	221.208	221.421	See	See

	estimator	function	n_samples_train	n_samples	n_features	mean_duration_sklearn	std_duration_sklearn	n_iter	iteration_throughput	latency	mean_duration_onnx	std_duration_onnx	accuracy_score_onnx	speedup	std_speedup	sklearn_profiling	onnx_profiling
1	HistGradientBoostingClassifier_best	predict	100000	1000	100	0.102	0.001	300	0.008	0.0	0.373	0.01	1.0	0.273	0.273	See	See